The Gamified Reading App Pilot

A biological framing treats dopamine as a neurotransmitter involved in reward prediction and motivated behaviour. Reward cues such as points, badges, and a leaderboard could plausibly engage dopamine-related circuits, so the claim is not unreasonable. But the headteacher is making a mechanism claim from a behavioural correlation. More logged minutes after badges does not show that dopamine caused the reading.

To move from plausible to supported, the school would need several kinds of evidence. First, temporal precedence: a dopamine-related proxy should change before or alongside the behaviour, not just after the whole programme. Second, dose-response: stronger reward cues should produce larger effects, while removing badges should reduce them. Third, manipulation: a controlled comparison, ideally random assignment, of badges versus a neutral reading log would test whether the reward feature is doing the work. Fourth, specificity: alternative explanations such as novelty, competition, teacher attention, or app logging errors must be ruled out.

Direct dopamine measurement with PET or fMRI is not feasible in a school and would be ethically disproportionate. A reasonable judgement is that the pilot supports, at most, a behavioural effect. The dopamine explanation remains plausible but unverified. The board should fund a controlled evaluation, not present the app as proof of a neural mechanism. This also guards against reductionism: a complex school behaviour is unlikely to be explained by one neurotransmitter alone.

The pilot has no control group, so maturation, novelty, teacher enthusiasm, and practice on the quiz are confounded with the app. Reading minutes are app-generated and vulnerable to students leaving the app open; the outcome is not independent of the intervention. The 20-item quiz may be narrow, and a pre-post gain could reflect teaching to the test rather than reading quantity. The cortisol component is sensible in timing because 8:30 a.m. reduces some diurnal variation, but two samples from 40 volunteers cannot establish a stress trajectory. Food, sleep, puberty, and anticipation of sampling all confound cortisol.

Two concrete improvements would help. First, randomly assign classes to the gamified app or a comparison reading log, keep the same quiz, and add an independent measure such as a timed reading task or teacher-observed reading. This tests whether the reward features, rather than extra attention, matter. Second, collect cortisol on two consecutive school days at the same time, with standard instructions about food and sleep, and add a self-report stress measure for triangulation. The school should pre-register primary outcomes and analyse by class to avoid contamination between conditions. These changes would make a re-run more informative without pretending the school can measure dopamine directly.

Cortisol sampling in minors is not automatically unethical, but it carries more risk than a questionnaire. It involves a biological specimen, can cause anxiety about health or stress, and may reveal information about a student's wellbeing. Consent must be voluntary for the student and include parents or guardians. Students should be able to opt out without academic penalty or teacher disapproval, and the school must avoid any implication that participation is linked to grades or favouritism.

Safeguards should include private collection, plain-language explanation of what cortisol measures and does not measure, secure storage, and anonymity in analysis. Individual results should not be fed back unless there is a clinically relevant finding and a clear support pathway. An ethics review, a data protection plan, and a named contact for questions are appropriate. The school should also consider whether a less intrusive measure, such as self-reported stress, would answer the same question. If the school cannot provide these safeguards, it should drop the cortisol component. The ethical judgement is that the biological measure must be proportionate to the question and respectful of the students, not merely convenient for the evaluation.